Three Paths from a Single Point: Why Girls Turn Pain Inward and Boys Turn It Outward

The article presents a synthesis of findings from neuroscience, criminology, and terrorism research, showing that self-harm, addiction, and radicalization share a common psychological mechanism — a response to adverse childhood experiences and the loss of personal significance. Gender does not determine the intensity of the pain, but rather the culturally conditioned channel through which it is expressed: girls turn pain inward (self-harm, depression), boys turn it outward (aggression, extremism). Kruglanski's 3N model (needs–narratives–networks) explains why these phenomena are interchangeable and points to the possibility of unified prevention instead of the current fragmented approach.

Self-harm, addiction, and extremism share a common psychological root. Data from neuroscience, criminology, and terrorism research show that gender does not determine the intensity of pain — only the channel through which it is expressed.

Imagine two fourteen-year-olds in the same classroom at a Czech primary school. Both have experienced the breakup of their family, bullying, a sense of invisibility. He began watching extremist content on Telegram. She began hurting herself. The teachers who noticed both problems referred them to two entirely different specialists — him to a radicalization-prevention program, her to a psychologist specializing in self-harm. Each received a different label, a different diagnosis, a different intervention.

And yet a new synthesis of data from neurobiology, addiction psychology, terrorism research, and criminology reveals something surprising: both are responding to the same pain through the same psychological mechanism — only by different means.

In the Czech Republic, an estimated more than 70,000 children aged 11–16 engage in repeated self-harm, predominantly girls (Váňová et al. 2024, Palacký University). Czech girls smoke considerably more than boys — 13.6% versus 9% in the 13–15 age group (GYTS 2022). Across Europe, on average only 49% of girls report good mental health compared with 69% of boys, and the Czech Republic, with an overall average of 46%, ranks among the worst in Europe (ESPAD 2024). Data from more than a hundred studies, databases on dozens of terrorist organizations, and clinical cohorts numbering tens of thousands of adolescents reveal that self-harm, smoking, and radicalization are not three distinct pathologies. They are three behavioral outputs of the same psychological program — and gender determines which output is activated.

Yet prevention of these phenomena is fragmented: self-harm is handled by school psychologists, addiction by addiction specialists, radicalization by security services. What if there were a form of prevention that protected against all three?

The answer begins in the anatomy of the adolescent brain. Between 2008 and 2010, Laurence Steinberg (Temple University) described what is known as the dual-systems model: the socioemotional-motivational system (the limbic system, the striatum) matures in early adolescence and sharply increases sensitivity to reward, excitement, and social cues. The cognitive control system (the prefrontal cortex, especially the dorsolateral prefrontal cortex), however, does not fully mature until around the age of 25. The result is a developmental imbalance: a powerful engine with inadequate brakes.

This imbalance creates a window of vulnerability common to all three risk behaviors. In smoking, it explains why adolescents underestimate the risk of addiction and overestimate the social reward of a "cool" identity. In self-harm, it explains the reduced ability to resist an impulse in a moment of intense emotion — Dahlgren, Hooley et al. (2018) demonstrated that adolescents who self-harm show significantly reduced activation of the dorsolateral prefrontal cortex, with the degree of reduction correlating with the degree of impulsivity. In radicalization, it explains the susceptibility to black-and-white narratives that offer a simple answer to a complex world.

An even deeper connection is revealed by the endogenous opioid system. Stanley et al. (2010, Journal of Psychiatric Research) found that individuals who self-harm have significantly lower baseline levels of beta-endorphin and met-enkephalin in their cerebrospinal fluid. Koenig et al. (2020, Neuropsychopharmacology), in an outpatient study of 51 women, confirmed that beta-endorphin levels are significantly lower before an act of self-harm than afterward, with more severe injuries associated with greater endorphin release. The mechanism is structurally identical to the way opiates compensate for an opioid deficiency in addicts. Nicotine activates the same mesolimbic dopamine system (ventral tegmental area → nucleus accumbens) by acting on nicotinic acetylcholine receptors. Blasco-Fontecilla et al. (2016, Frontiers in Psychiatry) identified the opioid and dopamine systems as the "most obvious targets" for interrupting the addictive cycle of self-harm — and proposed treatment with naltrexone and buprenorphine, drugs commonly used in opioid addiction.

It is no coincidence that all three pathways begin in the same place: with adverse childhood experiences (ACEs). The original study by Felitti et al. (1998, American Journal of Preventive Medicine) of more than 17,000 adults demonstrated a dose–response relationship: individuals with 4 or more ACEs are five times more likely to begin smoking early (Anda et al. 1999, JAMA) and 4–12 times more likely to develop alcohol or drug dependence. The population-attributable risk for illicit drug use is 67% (Dube et al. 2003) — two-thirds of illicit drug use is statistically attributable to adverse childhood experiences.

For self-harm, the gradient is even steeper. A systematic review in Alpha Psychiatry (2024) reports: 1–2 ACEs increase the likelihood of self-harm by a factor of 2.2, while 3 or more ACEs increase it sixfold. The strongest association is shown by emotional abuse (odds ratio 3.03), followed by sexual abuse (2.65) and physical violence (2.31) — according to the meta-analysis by Liu et al. (2018).

And for radicalization? Bélanger et al. validated the 3N model (needs–narratives–networks) in Canada, Pakistan, Spain, and the USA and identified social alienation as a key predictor. The clinical case study of a seventeen-year-old French girl, "Ley," convicted of ties to ISIS and documented in Frontiers in Psychiatry (Vandevoorde et al. 2022; cf. Rolling et al. 2022), revealed three generations of physical and sexual violence in the family, and a history of self-harm and suicide attempts prior to radicalization. The key finding: self-harm ceased at the moment Ley embraced a radical Islamist way of life — the radical group provided ritualization of daily life, support for self-image, and identification, exactly what self-harm had previously provided as a regulatory mechanism.

Johnson, McKernan, and Bruehl (2022) formulated a unified neurobiological mechanism: childhood trauma disrupts endogenous opioid function and programs the HPA axis (hypothalamic–pituitary–adrenal) — the same neurobiological changes that create vulnerability both to chronic pain and self-harm and to substance addiction.

The similarity between self-harm and smoking is not merely figurative — it is empirically documented. Nixon, Cloutier, and Aggarwal (2002, Journal of the American Academy of Child & Adolescent Psychiatry) adapted the DSM-IV criteria for substance dependence to fit self-harm in adolescent psychiatric patients. The results: 97% met three or more dependence criteria, 81% met five or more criteria. All participants reported an urge to self-harm following stressful events; 79% reported a near-daily urge (craving). A systematic review of 36 studies from 2000–2024 (Current Addiction Reports, Springer 2025) confirmed the addictive properties of deliberate self-harm, including compulsive urges, impaired control, and escalation of the behavior.

The comparison with nicotine addiction is instructive. Negative reinforcement: smoking to relieve withdrawal symptoms corresponds to self-harm to reduce unpleasant emotions (the primary function in 50–63% of episodes according to the meta-analysis by Taylor et al. 2018). Tolerance: the need for a higher dose corresponds to the need for more severe or more frequent injury. Craving: the physical and psychological craving for nicotine corresponds to the urge to self-harm (79% daily). Loss of control: continuation despite attempts to quit.

Direct evidence of co-occurrence comes from Moran et al. (2012, The Lancet, a cohort of 1,943 adolescents): newly emerging self-harm in adolescence is independently associated with cigarette smoking (hazard ratio 1.8), cannabis use (2.4), and risky alcohol drinking (2.1). A Finnish twin study (Korhonen et al.) adds precision: early initiation of smoking (before age 14) is associated with deliberate self-harm at age 22 (adjusted odds ratio 4.57). Specifically in girls, early smoking predicts suicidal ideation (3.69).

Particularly troubling is the "transfer effect": adolescents who have ceased repeated self-harm show markedly elevated rates of substance abuse (Kaess et al.). Without addressing the shared emotional dysregulation, one maladaptive strategy is simply replaced by another.

Here we reach the heart of the paradox. The data consistently show that girls exhibit higher rates of psychological pain than boys: roughly twice the incidence of depression (WHO, 2017; Salk et al. 2017, Psychological Bulletin), twice the rate of self-harm — in the youngest group, aged 10–14, by as much as an 8:1 ratio according to Czech clinical data (Rozsívalová, Paclt, and Trefilová, Česko-slovenská psychiatrie). In the Czech Republic, girls smoke considerably more than boys. And yet women make up only 10–15% of the membership of terrorist organizations worldwide (UN CTED, 2019).

The standard explanation — "women are less aggressive" — does not hold up on a closer look at the data. Geniole et al. (2020, Hormones and Behavior) conducted the most comprehensive meta-analysis to date of the testosterone–aggression relationship: the overall correlation was r = 0.054, in men r = 0.071, in women r = 0.002 (not significant). Pharmacological manipulation of testosterone showed even weaker and statistically non-significant causal effects (r = 0.046). Lise Eliot (2021, Current Anthropology) summarizes: girls learn faster than boys to suppress overt aggression, with gender awareness (emerging around age three) triggering a process of self-socialization.

A far more accurate explanation is offered by Achenbach's distinction between internalizing and externalizing. Internalizing problems (anxiety, depression, somatic complaints) are "problems within the self"; externalizing problems (rule-breaking, aggressive behavior) are "conflicts with others." An Indian study (2024) quantifies the gender difference: an average internalizing score of 18.6 in girls versus 7.6 in boys (t = 5.76, p < 0.01); an externalizing score of 28.93 in boys versus 14.33 in girls (t = 8.50, p < 0.01). Almost 87% of girls fell within the clinical range for internalizing.

Broidy and Agnew (1997, Justice Quarterly), in a seminal study, demonstrated that women experience the same or greater levels of strain as men, but their emotional response differs. Women's anger is accompanied by depression, anxiety, guilt, and shame (internalizing emotions), whereas men's anger manifests as moral outrage (an externalizing emotion). Men turn blame onto others — the result is outwardly directed deviance. Women turn blame inward — the result is inwardly directed deviance.

Nolen-Hoeksema (Yale) demonstrated, through response styles theory (1991), that women more often respond to negative mood with rumination (repeated brooding), whereas men are more likely to use distraction. And the key finding of Nolen-Hoeksema, Larson, and Grayson (1999): after statistically controlling for rumination, the gender difference in depression disappears — rumination statistically and fully mediates the gender gap in depression. This does not mean rumination is the sole cause, but that it represents the statistically strongest explanatory variable.

In Merton's theory of anomie (1938): men choose rebellion (active rejection of the existing system and its replacement, often through violence), women retreatism (withdrawal from both goals and means — chronic addictions, self-harm, psychological withdrawal). Not because women lack the capacity for rebellion, but because cultural narratives do not offer it to them.

Arie W. Kruglanski (University of Maryland) developed Significance Quest Theory (SQT) and the 3N model of radicalization (Kruglanski et al. 2014, Political Psychology; Kruglanski, Bélanger, and Gunaratna 2019, The Three Pillars of Radicalization, Oxford University Press). The model identifies three elements that produce radicalization: the need for personal significance, a narrative justifying the means to significance, and a network that affirms the narrative.

Although Kruglanski originally developed the model to explain violent extremism, his expanded work Psychology of the Extreme (with Moskalenko) explicitly includes addiction, gambling, internet addiction, and self-destructive behavior as manifestations of the same psychological mechanism — a motivational imbalance in which one dominant need suppresses all others. In 2022 (Perspectives on Psychological Science), he extended SQT as a general theory of extreme behavior.

Applying the 3N model to all three phenomena reveals a symmetry:

In smoking, the need is social exclusion, stress, "not being cool." The narrative is "smoking is grown-up, rebellious, calming." The network is the peer group of smokers that normalizes the behavior.

In self-harm, the need is emotional pain, rejection, invisibility. The narrative is "hurting myself helps, it gives me control." The network is the self-harm community on social media that affirms the behavior.

In radicalization, the need is humiliation, social exclusion, loss of significance. The narrative is "violence will restore honor and justice." The network is the extremist group that legitimizes the behavior.

The need is identical in all three cases. The key differentiating factor is the narrative — and the narrative is gender-coded. Cultural narratives systematically fail to offer violence as a legitimate path to significance for women. The hero's narrative, the martyr's narrative, the warrior's narrative — all of these are culturally coded as male.

Ellenberg, Speckhard, and Kruglanski (2023, Psychology of Men and Masculinities, APA) demonstrated this mechanism directly in a study of the incel community (N = 272). They identified three distinct narrative-driven responses to an identical loss of significance (romantic/social rejection): an externalizing narrative promoting violence (radicalization), an internalizing narrative promoting suicide/self-harm, and a hopeful narrative leading to constructive adaptation. Both externalizing and internalizing share the same unmet need; their behavioral divergence is exclusively the result of narrative framing.

Contu, Ellenberg, Kruglanski, and Pierro (2023, Frontiers in Psychology) conducted the first direct empirical test of the substitutability of means: loss of significance in the romantic domain predicts extremism in the professional domain, and vice versa. When one channel for restoring significance is blocked, psychological energy is redirected into another.

The most compelling evidence that women's low participation in terrorism is a consequence of cultural barriers — not a biological incapacity for violence — comes from groups with egalitarian ideologies.

Wood and Thomas (2017, Journal of Peace Research) analyzed a dataset of more than 400 insurgent organizations active between 1979 and 2009 and demonstrated that a Marxist orientation statistically significantly increases the proportion of female fighters, whereas Islamist ideology decreases it. There is an inverse relationship between a group's religiosity and the representation of women.

The German Red Army Faction (RAF) represents an extreme case: women made up roughly 50% of members and as much as 80% of sympathizers (MacDonald, Shoot the Women First, 1991). Not only did they participate — they led. Ulrike Meinhof and Gudrun Ensslin were co-founders. Brigitte Mohnhaupt directed the "German Autumn" campaign of 1977.

The Kurdish PKK offers a model in which feminist ideology is directly encoded into the organizational structure. Women make up roughly 40% of the fighting force, and in the 1990s they carried out 11 of the PKK's 15 suicide attacks in Turkey (73%), despite constituting only 22% of operatives. The all-female YPJ units played a key role in the fight against ISIS.

The LTTE systematically recruited women — they made up roughly 30% of the total force and 24–40% of suicide attackers. Twenty-two-year-old Kalaivani Rajaratnam (known as Thenmozhi or Dhanu) assassinated Indian Prime Minister Rajiv Gandhi in 1991.

Boko Haram deployed more than 468 women and girls in 240 suicide attacks, with women making up more than 50% of all its bombers (Warner and Matfess 2017, CTC West Point) — here, however, it was predominantly a matter of coercion rather than voluntariness. The youngest bombers were seven years old.

Europol data show a fluctuating trend: the proportion of women among terrorism-related arrests in the EU rose from 18% in 2015 to 26% in 2016, then fell to 9.6% in 2024 (TE-SAT 2025).

The evidence is consistent: where ideology removes gender barriers, women's participation rises sharply. The problem is not primarily biological — it is cultural.

The Czech Republic exhibits a unique combination of risk factors. Czech girls smoke considerably more than boys: 13.6% versus 9% (GYTS 2022). ESPAD 2024 reveals a dramatic gender gap in mental health: across Europe, on average only 49% of girls report good mental health versus 69% of boys — and the Czech Republic, with an overall average of 46%, ranks below the Europe-wide average of 59%.

The longitudinal study by Váňová et al. (2024, Palacký University, N = 1,299, three years) found a self-harm prevalence of 39.8–44.09% across waves, with 9.72% exhibiting high-frequency deliberate self-harm across all three waves. A pilot survey by the National Institute of Mental Health (NÚDZ) (6,000 ninth-grade pupils) found reduced well-being in 50%, depressive symptoms in 40%, and elevated anxiety in 30%. The Safety Line (Linka bezpečí) recorded 4,725 calls about self-harm in 2023 — nearly double the figure for 2022.

The Czech system faces serious capacity constraints. A shortage of child psychiatrists and psychologists means waiting times running into months. Suicide has become the second most common cause of death among Czech adolescents. And yet programs for the prevention of self-harm, addiction, and radicalization are managed separately — by different ministries, different specialists, with different budgets. Self-harm falls under the Ministry of Education and the health sector, addiction under the government (the national anti-drug coordinator) and addiction services, radicalization under the Ministry of the Interior. Each sector has its own prevention strategy, its own methodologies, its own funding — and none systematically addresses the shared roots.

The thesis of the functional equivalence of self-harm, addiction, and radicalization has its limits and deserves an honest confrontation with counterarguments.

First — self-harm and terrorism differ dramatically in their consequences. Self-harm primarily harms the individual; terrorism kills the innocent. Functional equivalence at the psychological level does not imply moral equivalence. Critics rightly object that merging the two phenomena into a single category may trivialize terrorism or, conversely, stigmatize self-harm.

Second — the biological contribution to gender differences in aggression is not zero, even if it is weaker than traditionally assumed. The testosterone–aggression correlation of r = 0.054 is low, but statistically significant in men. Prenatal exposure to androgens influences brain development in ways that are not fully offset by socialization. The "it's all socialization" model is an oversimplification — the reality lies between biological determinism and pure social constructivism.

Third — the substitutability of means is not complete. Victor, Glenn, and Klonsky (2012, Psychiatry Research) identified a key difference: self-harm is maintained almost exclusively by negative reinforcement (emotion regulation), whereas substances also involve significant positive reinforcement (inducing pleasurable states). Self-harm is a "purer" form of emotion regulation than classic addiction — and radicalization additionally offers a social identity and group belonging that self-harm does not provide.

Fourth — the 3N model has been validated cross-culturally (Canada, Pakistan, Spain, USA), but a systematic test of its application to gender differences in self-harm has not yet been conducted. Extrapolating from radicalization research to self-harm is theoretically consistent but empirically still incomplete.

Fifth — most of the cited studies on self-harm and smoking have a cross-sectional design, which does not permit causal conclusions. Longitudinal data (Váňová et al., Moran et al.) are more robust, but even these are subject to limitations — they measure associations, not necessarily causal relationships. ACE studies are retrospective and subject to recall bias.

What remains valid, however, even after accounting for these counterarguments: the neurobiological pathways (the opioid system, the dopamine system, the HPA axis, the immature prefrontal cortex) are shared. The risk factors (ACEs, loneliness, bullying, social exclusion) are shared. The primary function (regulation of an unbearable state of mind) is shared. And where cultural barriers fall, the gender difference in violent behavior shrinks dramatically.

If self-harm, addiction, and radicalization share a common psychological root, then there should exist an intervention that protects against all three at once. And there are programs with relevant evidence.

Dialectical behavior therapy for adolescents (DBT-A) has the strongest evidence base for reducing both self-harm and suicide attempts. A meta-analysis in European Child & Adolescent Psychiatry (2025) quantifies it: DBT-A reduces the proportion of adolescents who self-harm (risk difference −0.12), suicidal ideation (mean difference −9.8), and specifically deliberate self-harm (risk difference −0.20). The DBT modules — mindfulness, distress tolerance, emotion regulation, interpersonal effectiveness — address exactly the deficits that are shared with vulnerability to both addiction and radicalization.

Gilbert Botvin's Life Skills Training (LST) program, originally developed for substance-use prevention, effectively strengthens emotional regulation and coping skills — precisely what also protects against self-harm and radicalization.

Youth Aware of Mental Health (YAM) is a universal school-based suicide-prevention program for adolescents aged 13–17. In the SEYLE study (Wasserman et al. 2015; approximately 11,000 ninth-grade pupils across ten European countries), YAM was the only intervention that significantly reduced suicidal outcomes: 55% fewer suicide attempts and 50% fewer cases of severe suicidal ideation over a one-year horizon. NNT = 91: for every 91 students enrolled, one severe case is prevented.

Shared protective factors confirmed across all three phenomena include family connectedness and communication, school belonging (a positive school climate mitigates risks — the #BeeWell study, N = 12,130), emotion-regulation skills (the most consistent protective factor), and meaningful activities that provide a sense of personal significance. A systematic review (MDPI 2025) explicitly recommends the integrated screening of emotional symptomatology, self-regulation strategies, and substance use for the early identification of at-risk adolescents.

In practice, for the Czech Republic this would mean: a single integrated prevention program in schools that jointly addresses emotional regulation, healthy coping strategies, media literacy (against algorithmic escalation), school belonging, and family communication — instead of three separate programs for three separate ministries.

Let us return to those two fourteen-year-olds. He on Telegram, she with scars on her soul and on her body. Different symptoms, different specialists, different programs. And yet the same chronically elevated cortisol, the same deficient opioid system, the same immature prefrontal cortex, the same ACE score, the same sense of invisibility. The same feeling that "no one cares about me" — what Gordon Flett (York University, 2022) calls anti-mattering, the sense of one's own insignificance, and what Kruglanski calls the loss of personal significance.

The difference between them was not determined by biology. It was determined by the narrative their culture offered them, and the network that took them in. To him, the online community said: "Your anger is justified, the enemy is clear." To her, the online community said: "Cutting helps, you're in this with us." Both found temporary relief — and both fell into a spiral of escalation.

A single integrated program focused on emotional regulation, school belonging, family communication, and a sense of personal significance could protect both of them. Not because self-harm and radicalization are the same thing. But because they grow from the same root — and the most effective prevention targets the root, not the branches.

This analysis is based on a synthesis of peer-reviewed studies, institutional reports (UN CTED, Europol TE-SAT, WHO, START), and databases of terrorist organizations; an AI assistant was used in its preparation. Data cutoff: February 2026. Key sources: Kruglanski et al. 2014, 2022; Steinberg 2008, 2010; Nixon, Cloutier & Aggarwal 2002; Wood & Thomas 2017; Ellenberg, Speckhard & Kruglanski 2023; Broidy & Agnew 1997; Bloom 2011; Felitti et al. 1998. We recommend independent verification before using it for decision-making.

Transparency of Creation

The conception, structure, and editorial line of the article are the work of the author, who prepared the content outline, established the key theses, and directed the entire creative process. Generative AI (Claude Opus 4.6, Anthropic) was used as a tool for research, fact-checking, and fleshing out the author's draft.

The author verified the key findings and approved the final wording. No part of the text was published without conscious authorial oversight. Factual data were verified against the publicly available sources cited in the text.

This procedure complies with the transparency principles of EU Regulation 2024/1689 (the AI Act). #poweredByAI